[1]邵怡蕾.智能盈余的再生产：当五位诺奖得主的假设同时失效[J].清华大学教育研究,2026,(02):24-35.[doi:10.14138/j.1001-4519.2026.02.002411]
SHAO Yi-lei.The Reproduction of Intelligent Surplus:When Five Nobel-Prize Assumptions Fail at Once[J].TSINGHUA JOURNAL OF EDUCATION,2026,(02):24-35.[doi:10.14138/j.1001-4519.2026.02.002411]
点击复制
智能盈余的再生产：当五位诺奖得主的假设同时失效
SHAO Yi-lei.The Reproduction of Intelligent Surplus:When Five Nobel-Prize Assumptions Fail at Once[J].TSINGHUA JOURNAL OF EDUCATION,2026,(02):24-35.[doi:10.14138/j.1001-4519.2026.02.002411]
清华大学教育研究[ISSN:1001-4519/CN:11-1610/G4]
- 卷:
- 期数:
- 2026年02期
- 页码:
- 24-35
- 栏目:
- 教育前沿
- 出版日期:
- 2026-04-20
文章信息/Info
- Title:
- The Reproduction of Intelligent Surplus:When Five Nobel-Prize Assumptions Fail at Once
- Keywords:
- Intelligent Surplus; Silicon-Based Economics; dual literacy; First Compressor; meaningtranslation capacity; intelligent infrastructure
- 分类号:
- G40-05
- 文献标志码:
- A
- 摘要:
- 当大语言模型可以即时生成专业文本并替代大量认知劳动，“人力资本”这个支撑全球教育政策七十年的核心概念还成立吗？本文回溯舒尔茨、贝克尔、斯宾塞、罗默、卢卡斯五位诺奖得主的理论遗产，提炼出教育经济学五大核心假设——线性回报、稳定折旧、个体归属、稳定信号、单向外部性——并论证它们在AI时代的同时失效是由于同一根源：智能本身不再稀缺。本文提出“智能盈余”替代“人力资本”作为教育经济学新的核心分析单位，借鉴布尔迪厄文化资本框架将其区分为体化（人的素养）、物化（技术系统）、制度化（治理框架）三种形态。进而追问：当智能不再稀缺，什么才是真正稀缺的？AI系统以压缩为核心逻辑但趋向同质化和无意义化，唯有能将独特的人类经验首次翻译为AI系统可理解表征的“第一压缩者”，才能为AI系统注入不可替代的输入。这种“意义翻译能力”来自对具体他者的真实在意，是阿伦特意义上“行动”的前提。由此，教育的社会功能从“人力资本的生产”转向“智能盈余的再生产”——即在个体层面培养技术素养与人文素养不可分离的“双元素养”，在组织层面为意义翻译能力的涌现创造条件，在国家层面将教育视为智能文明的基础设施。教育的终极目标，从培养劳动者转向培养"城邦中的行动者"——那些在人机共生时代能够进行意义翻译的人。
- Abstract:
- When large language models can instantly generate professional text and displace cognitive labor at scale, does “human capital”—the concept anchoring global education policy for seventy years—still hold?This paper revisits the theoretical legacy of ?ve Nobel laureates (Schultz, Becker, Spence, Romer, Lucas) and distills five core assumptions of the economics of education: linear returns, stable depreciation, individual attribution, stable signaling, and unidirectional externalities. It argues that all ?ve are failing simultaneously,driven by a single cause: intelligence itself is no longer scarce. The paper proposes “Intelligent Surplus” to replace “human capital” as the field’s core analytical unit, distinguishing three forms: embodied (human competence), objecti?ed (technical systems), and institutionalized (governance frameworks). It further asks: if intelligence is no longer scarce, what truly is? AI systems are grounded in compression as their core logic, yet they tend toward homogeneity and meaninglessness. Only the “First Compressors,” who can translate unique human experiences into intelligible representations for AI systems for the ?rst time, are capable of providing irreplaceable input for such systems. This “capacity for meaning translation” arises from genuine concern for concrete others—the precondition for “action” in the Arendtian sense. Accordingly, the social function of education shifts from the production of human capital to the reproduction of intelligent surplus: at the individual level, it fosters “dual literacy” that integrates technological literacy and humanistic literacy inseparably; at the organizational level, it creates conditions for the emergence of meaning-translation capacity; and at the national level, it regards education as the infrastructure of an intelligent civilization. The ultimate goal of education is thus transformed from cultivating laborers to nurturing “actors in the polis”—individuals capable of meaning translation in the era of human-AI symbiosis.
更新日期/Last Update:
2026-04-20